铁路路基层面的检测与跟踪

Detecting and Tracing of Railroad Trackbed Interfaces

利用小波变换将原始雷达记录变换为层面突变信息占主要成分的信号,根据电磁波在层面位置反射回波的特征,设计出层面检测算法,提取出所有道数据的层面点信息,针对层面出现的断裂、跳跃、分岔、交叉等现象,提出层面点归类原则,从而智能地实现层面的跟踪.通过实际资料验证表明,该算法能快速地绘制出铁路路基的层面曲线图,并能反映地下结构层的真实情况,适用于海量数据处理,将技术人员从繁重的读图任务中解脱出来.

Based on the wavelet transform method, the primitive Ground Penetrating Radar （GPR） data were transformed into the signal whose principal constituent was only the sudden changes of the interface information. An interface-test algorithm strategy is designed according to the characteristic of reflection echo from interfaces. By extracting the information of individual points of the interfaces from all traced data,an interface classification principle is proposed to distinguish the rupture, fluctuation, branching and intersection of the interface, so that the interfaces are intelligently traced. The actual testing shows that the method can be used to plot the graph of railroad trackbed interfaces swiftly, reflect the actual conditions of railroad substructures, and process a mass of data. The engineering technicians are released from the burden of the image interpretation.